Self-Adaptive Ageing Models for Optimal Management and Planning of Assets in Microgrids

The evolution towards decentralization of the electricity grids leads towards a large increase of new and flexible devices in low- and medium-voltage smart grid environments such as microgrids and energy communities. In this type of environment, an advanced, dynamic and autonomous asset management system will be essential to ensure grid optimal performance. Self-adaptive ageing modelling algorithms using real-life historical operational data of assets to assess the future capacity, remaining useful life and optimal operational profiles are essential tools in such a context. This paper will propose a modelling methodology for this purpose, based on machine learning techniques using historical data of assets operated in real-life circumstances and with a limited knowledge of asset properties. The process flow of the methodology is explained in detail, as well as the tangible outputs that are generated. The method was applied on a photovoltaic system in a real-life environment and first experimental results are discussed.